Video data inter prediction based on control points
A video data inter prediction method is provided, which includes: determining a candidate motion information list of a current picture block, where the candidate motion information list includes at least one first candidate motion information group, at least one second candidate motion information group, the first candidate motion information group is a motion information group determined based on motion information of preset locations on a first neighboring picture block of the current picture block and a motion model of the first neighboring picture block, the second candidate motion information group is a set of motion information of at least two sample locations that are respectively neighboring to at least two preset locations on the current picture block; determining target motion information from the candidate motion information list; and performing inter prediction on the current picture block based on the target motion information.
1. A method, comprising:
determining a candidate motion information list of a current picture block to be coded using an affine merge mode, the candidate motion information list comprising at least one first candidate motion information group and at least one second candidate motion information group both comprising motion vectors for a first group of at least two control points of the current picture block, wherein a plurality of pieces of index information indicate the first candidate motion information group and the second candidate motion information group respectively, the first candidate motion information group is a motion information group determined based on motion information of control points of a first affine neighboring picture block of the current picture block and a motion model of the first affine neighboring picture block, and the second candidate motion information group is based on a set of motion information of at least two sample locations that are respectively neighboring to control points of a second group of at least two control points of the current picture block, and the at least two sample locations are located on at least one second neighboring picture block of the current picture block;
determining target motion information from the candidate motion information list; and
performing inter prediction on the current picture block based on the target motion information; and
wherein the determining the candidate motion information list of the current picture block comprises:
in response to that the second group of at least two control points is different from the first group of at least two control points, deriving the second candidate motion information group according to a location transformation formula and based on motion information corresponding to the second group of at least two control points of the current picture block.
2. The method according to claim 1 , wherein determining the candidate motion information list of the current picture block comprises:
in response to that the first affine neighboring picture block is a picture block using an affine motion model, deriving, based on motion information of at least two control points of the first affine neighboring picture block and the affine motion model of the first affine neighboring picture block, motion information of at least two control points of the current picture block, and adding the motion information of the at least two control points of the current picture block into the candidate motion information list as the first candidate motion information group, wherein the at least two control points of the first affine neighboring picture block corresponds to the at least two control points of the current picture block.
3. The method according to claim 1 , wherein an affine motion model of the current picture block and an affine motion model of the first affine neighboring picture block are four-parameter affine motion models; or
an affine motion model of the current picture block and an affine motion model of the first affine neighboring picture block are six-parameter affine motion models.
4. The method according to claim 1 , wherein determining the candidate motion information list of the current picture block comprises:
adding the at least one first candidate motion information group to the candidate motion information list, and subsequently adding the at least one second candidate motion information group to the candidate motion information list, and
adding zero motion information to the candidate motion information list in response to that a length of the candidate motion information list is less than a predefined length N upon having added the at least one first candidate motion information group and the at least one second candidate motion information group.
5. The method according to claim 1 , wherein the location transformation formula is based on a 4-parameter affine model or a 6-parameter model and coordinates of the first group of control points and the second group of control points.
6. The method according to claim 1 , wherein:
the second group of at least two control points is a top-left control point, a top-right control point, and a bottom-right control point of the current picture block and the first group of at least two control points is the top-left control point, the top-right control point, and a bottom-left control point of the current picture block; or
the second group of at least two control points is a top-left control point, a bottom-left control point, and a bottom-right control point of the current picture block and the first group of at least two control points is the top-left control point, a top-right control point, and the bottom-left control point of the current picture block; or
the second group of at least two control points is a top-right control point, a bottom-left control point, and a bottom-right control point of the current picture block and the first group of at least two control points is a top-left control point, the top-right control point, and the bottom-left control point of the current picture block; or
the second group of at least two control points is a top-left control point and a bottom-left control point of the current picture block and the first group of at least two control points is the top-left control point and a top-right control point of the current picture block.
7. A device, comprising:
one or more processors; and
a non-transitory computer-readable storage medium coupled to the one or more processors and storing programming for execution by the one or more processors, wherein the programming, when executed by the one or more processors, causes the device to:
determine a candidate motion information list of a current picture block to be coded using an affine merge mode, wherein the candidate motion information list comprises at least one first candidate motion information group and at least one second candidate motion information group both comprising motion vectors for a first group of at least two control points of the current picture block, wherein a plurality of pieces of index information indicate the first candidate motion information group and the second candidate motion information group respectively, the first candidate motion information group is a motion information group determined based on motion information of control points of a first affine neighboring picture block of the current picture block and a motion model of the first affine neighboring picture block, and the second candidate motion information group is based on a set of motion information of at least two sample locations that are respectively neighboring to control points of a second group of at least two control points of the current picture block, and the at least two sample locations are located on at least one second neighboring picture block of the current picture block;
determine target motion information from the candidate motion information list; and
perform inter prediction on the current picture block based on the target motion information; and
wherein the determining the candidate motion information list of the current picture block comprises:
in response to that the second group of at least two control points is different from the first group of at least two control points, deriving the second candidate motion information group according to a location transformation formula and based on motion information corresponding to the second group of at least two control points.
8. The device according to claim 7 , wherein the programming, when executed by the one or more processors, causes the device to determine the candidate motion information list of the current picture block by:
in response to that the first affine neighboring picture block is a picture block using an affine motion model, deriving, based on motion information of at least two control points of the first affine neighboring picture block and the affine motion model of the first affine neighboring picture block, motion information of at least two control points of the current picture block, and adding the motion information of the at least two control points of the current picture block into the candidate motion information list as the first candidate motion information group, wherein the at least two control points of the first affine neighboring picture block corresponds to the at least two control points of the current picture block.
9. The device according to claim 7 , wherein an affine motion model of the current picture block and an affine motion model of the first affine neighboring picture block are four-parameter affine motion models; or
an affine motion model of the current picture block and an affine motion model of the first affine neighboring picture block are six-parameter affine motion models.
10. The device according to claim 7 , wherein the programming, when executed by the one or more processors, causes the device to determine the candidate motion information list of the current picture block by:
adding the at least one first candidate motion information group to the candidate motion information list, and subsequently adding the at least one second candidate motion information group to the candidate motion information list; and
adding zero motion information to the candidate motion information list in response to that a length of the candidate motion information list is less than a predefined length N upon having added the at least one first candidate motion information group and the at least one second candidate motion information group.
11. The device according to claim 7 , wherein the location transformation formula is based on a 4-parameter affine model or a 6-parameter model and coordinates of the first group of control points and the second group of control points.
12. The device according to claim 7 , wherein:
the second group of at least two control points is a top-left control point, a top-right control point, and a bottom-right control point of the current picture block and the first group of at least two control points is the top-left control point, the top-right control point, and a bottom-left control point of the current picture block; or
the second group of at least two control points is a top-left control point, a bottom-left control point, and a bottom-right control point of the current picture block and the first group of at least two control points is the top-left control point, a top-right control point, and the bottom-left control point of the current picture block; or
the second group of at least two control points is a top-right control point, a bottom-left control point, and a bottom-right control point of the current picture block and the first group of at least two control points is a top-left control point, the top-right control point, and the bottom-left control point of the current picture block; or
the second group of at least two control points is a top-left control point and a bottom-left control point of the current picture block and the first group of at least two control points is the top-left control point and a top-right control point of the current picture block.
13. A non-transitory computer-readable medium carrying a program code which, when executed by a computer device, causes the computer device to perform steps comprising:
determining a candidate motion information list of a current picture block to be coded using an affine merge mode, the candidate motion information list comprising at least one first candidate motion information group and at least one second candidate motion information group both comprising motion vectors for a first group of at least two control points of the current picture block, wherein a plurality of pieces of index information indicate the first candidate motion information group and the second candidate motion information group respectively, the first candidate motion information group is a motion information group determined based on motion information of control points of a first affine neighboring picture block of the current picture block and a motion model of the first affine neighboring picture block, and the second candidate motion information group is based on a set of motion information of at least two sample locations that are respectively neighboring to control points of a second group of at least two control points of the current picture block, and the at least two sample locations are located on at least one second neighboring picture block of the current picture block;
determining target motion information from the candidate motion information list; and
performing inter prediction on the current picture block based on the target motion information; and
wherein the determining the candidate motion information list of the current picture block comprises:
in response to that the second group of at least two control points is different from the first group of at least two control points, deriving the second candidate motion information group according to a location transformation formula and based on motion information corresponding to the second group of at least two control points.
14. The non-transitory computer-readable medium according to claim 13 , wherein the determining the candidate motion information list of the current picture block comprises:
when the first affine neighboring picture block is a picture block using an affine motion model, deriving, based on motion information of at least two control points of the first affine neighboring picture block and the affine motion model of the first affine neighboring picture block, motion information of at least two control points of the current picture block, and adding the motion information of the at least two control points of the current picture block into the candidate motion information list as the first candidate motion information group, wherein the at least two control points of the first affine neighboring picture block corresponds to the at least two control points of the current picture block.
15. The non-transitory computer-readable medium according to claim 13 , wherein an affine motion model of the current picture block and an affine motion model of the first affine neighboring picture block are four-parameter affine motion models; or
an affine motion model of the current picture block and an affine motion model of the first affine neighboring picture block are six-parameter affine motion models.
16. The non-transitory computer-readable medium according to claim 13 , wherein determining the candidate motion information list of the current picture block comprises:
adding the at least one first candidate motion information group to the candidate motion information list, and subsequently adding the at least one second candidate motion information group to the candidate motion information list, and
adding zero motion information to the candidate motion information list in response to that a length of the candidate motion information list is less than a predefined length N upon having added the at least one first candidate motion information group and the at least one second candidate motion information group.
17. The non-transitory computer-readable medium according to claim 13 , wherein the location transformation formula is based on a 4-parameter affine model or a 6-parameter model and coordinates of the first group of control points and the second group of control points.
18. The non-transitory computer-readable medium according to claim 13 , wherein:
the second group of at least two control points is a top-left control point, a top-right control point, and a bottom-right control point of the current picture block and the first group of at least two control points is the top-left control point, the top-right control point, and a bottom-left control point of the current picture block; or
the second group of at least two control points is a top-left control point, a bottom-left control point, and a bottom-right control point of the current picture block and the first group of at least two control points is the top-left control point, a top-right control point, and the bottom-left control point of the current picture block; or
the second group of at least two control points is a top-right control point, a bottom-left control point, and a bottom-right control point of the current picture block and the first group of at least two control points is a top-left control point, the top-right control point, and the bottom-left control point of the current picture block; or
the second group of at least two control points is a top-left control point and a bottom-left control point of the current picture block and the first group of at least two control points is the top-left control point and a top-right control point of the current picture block.